Evidence before optimization

See what AI systems and homeowners can actually understand about your business.

The AI Recommendation Audit turns public evidence, local relevance, site knowledge, and technical clarity into a cited diagnostic. It identifies what is supported, what is missing, and what to improve next.

Request an audit discussion Review the framework

Five-dimension framework

A 100-point diagnostic with evidence attached

Each dimension measures observable readiness. The total is not a prediction that Google, ChatGPT, Gemini, Copilot, Perplexity, or Claude will recommend a business.

20 points

Entity authority

Can systems identify the business and corroborate who it is?

Published scoring criteria

Canonical identity and structured entity: 10 points
Consistent corroborating references: 10 points

Evidence reviewed

Canonical identity and structured organization data
Consistent public business facts
Supported third-party references and identifiers

25 points

Trust signals

Which trust claims are visible, current, sourced, and specific?

Published scoring criteria

Dated, source-labeled trust evidence: 15 points
Visible claim and sponsorship boundaries: 10 points

Evidence reviewed

Ownership, license, insurance, and freshness states
Review-source and complaint-resolution evidence
Clear separation of supplied, checked, and observed facts

20 points

Knowledge coverage

Does the site answer the homeowner questions surrounding the service?

Published scoring criteria

Service and problem knowledge coverage: 10 points
Sources, reviewers, and freshness: 10 points

Evidence reviewed

Service, emergency, cost, maintenance, and selection guidance
Visible sources, reviewer disclosure, and review dates
Useful answers beyond commodity business descriptions

20 points

Local authority

Is service-area relevance supported rather than repeated?

Published scoring criteria

Specific service-area evidence: 10 points
Locally differentiated resources: 10 points

Evidence reviewed

Specific cities and services actually covered
Local resources, constraints, and problem context
Corroborating local relationships and references

15 points

AI readability

Can crawlers retrieve and correctly interpret the evidence?

Published scoring criteria

Crawl, index, canonical, and sitemap readiness: 8 points
Visible content and schema alignment: 7 points

Evidence reviewed

Indexability, canonical, sitemap, and crawler access
Visible content aligned with accurate structured data
Clear headings, internal links, citations, and entity relationships

What you receive

A blueprint tied to findings, not generic AI advice.

Evidence inventory with supplied, independently checked, observed, and missing states
Five-dimension score with cited findings and no invented facts
Search and AI-surface observation table separating mentions, links, and citations
Prioritized 30-day blueprint tied to pages, evidence, and measurable acceptance gates
Plain-language limitations explaining what the audit cannot promise

Non-negotiable boundaries

No promise to make an AI system recommend you

Recommendation systems change, personalize results, use different source sets, and remain outside THP control.

The score measures evidence and technical readiness; it is not the probability that an AI assistant will recommend a business.
A high score does not guarantee rankings, citations, leads, bookings, or revenue.
Blocked or sign-in-gated surfaces remain unobserved and are not replaced with inferred results.
Payment, directory inclusion, profile ownership, and sponsorship do not increase trust evidence scores.
Missing evidence scores as missing; contractor-supplied claims are not relabeled as independently verified.

Common questions

What the audit can—and cannot—establish

Does the audit guarantee an AI recommendation?

No. The audit measures observable evidence and technical readiness. It does not guarantee rankings, citations, recommendations, leads, bookings, or revenue.

Can payment improve the trust score?

No. Payment, sponsorship, directory inclusion, and profile ownership do not increase trust evidence scores.

How are missing or blocked sources scored?

Missing and not-observable criteria receive zero points and remain labeled. They are never removed from the denominator or inferred from another surface.

Start with evidence

Request a scoped audit discussion.

THP will confirm the business, service area, available sources, and observation scope before representing any score as complete.

Request an audit discussion

Evidence before optimization

See what AI systems and homeowners can actually understand about your business.

Request an audit discussion Review the framework

Five-dimension framework

A 100-point diagnostic with evidence attached

Each dimension measures observable readiness. The total is not a prediction that Google, ChatGPT, Gemini, Copilot, Perplexity, or Claude will recommend a business.

20 points

Entity authority

Can systems identify the business and corroborate who it is?

Published scoring criteria

Canonical identity and structured entity: 10 points
Consistent corroborating references: 10 points

Evidence reviewed

Canonical identity and structured organization data
Consistent public business facts
Supported third-party references and identifiers

25 points

Trust signals

Which trust claims are visible, current, sourced, and specific?

Published scoring criteria

Dated, source-labeled trust evidence: 15 points
Visible claim and sponsorship boundaries: 10 points

Evidence reviewed

Ownership, license, insurance, and freshness states
Review-source and complaint-resolution evidence
Clear separation of supplied, checked, and observed facts

20 points

Knowledge coverage

Does the site answer the homeowner questions surrounding the service?

Published scoring criteria

Service and problem knowledge coverage: 10 points
Sources, reviewers, and freshness: 10 points

Evidence reviewed

Service, emergency, cost, maintenance, and selection guidance
Visible sources, reviewer disclosure, and review dates
Useful answers beyond commodity business descriptions

20 points

Local authority

Is service-area relevance supported rather than repeated?

Published scoring criteria

Specific service-area evidence: 10 points
Locally differentiated resources: 10 points

Evidence reviewed

Specific cities and services actually covered
Local resources, constraints, and problem context
Corroborating local relationships and references

15 points

AI readability

Can crawlers retrieve and correctly interpret the evidence?

Published scoring criteria

Crawl, index, canonical, and sitemap readiness: 8 points
Visible content and schema alignment: 7 points

Evidence reviewed

Indexability, canonical, sitemap, and crawler access
Visible content aligned with accurate structured data
Clear headings, internal links, citations, and entity relationships

What you receive

A blueprint tied to findings, not generic AI advice.

Evidence inventory with supplied, independently checked, observed, and missing states
Five-dimension score with cited findings and no invented facts
Search and AI-surface observation table separating mentions, links, and citations
Prioritized 30-day blueprint tied to pages, evidence, and measurable acceptance gates
Plain-language limitations explaining what the audit cannot promise

Non-negotiable boundaries

No promise to make an AI system recommend you

Recommendation systems change, personalize results, use different source sets, and remain outside THP control.

The score measures evidence and technical readiness; it is not the probability that an AI assistant will recommend a business.
A high score does not guarantee rankings, citations, leads, bookings, or revenue.
Blocked or sign-in-gated surfaces remain unobserved and are not replaced with inferred results.
Payment, directory inclusion, profile ownership, and sponsorship do not increase trust evidence scores.
Missing evidence scores as missing; contractor-supplied claims are not relabeled as independently verified.

Common questions

What the audit can—and cannot—establish

Does the audit guarantee an AI recommendation?

No. The audit measures observable evidence and technical readiness. It does not guarantee rankings, citations, recommendations, leads, bookings, or revenue.

Can payment improve the trust score?

No. Payment, sponsorship, directory inclusion, and profile ownership do not increase trust evidence scores.

How are missing or blocked sources scored?

Missing and not-observable criteria receive zero points and remain labeled. They are never removed from the denominator or inferred from another surface.

Start with evidence

Request a scoped audit discussion.

THP will confirm the business, service area, available sources, and observation scope before representing any score as complete.

Request an audit discussion